Dark matter remains one of the most significant scientific mysteries of our time. This Resource Letter provides an overview of the astrophysical evidence for dark matter and describes the broad set of theoretical dark matter candidates that have been proposed. Results from dark matter searches are discussed, with a focus on direct detection experiments.

All known particles—electrons, protons, and neutrons (and their constituent quarks), along with the rest of the Standard Model of particle physics—compose just 15% of the matter budget of the universe. The balance is dark matter (DM)—invisible matter that interacts gravitationally with the rest of the universe but does not interact with light. Uncovering the physics of DM has proven an elusive challenge, and currently stands as one of the biggest open questions in physics.

Early evidence for invisible matter came in 1933 when astronomer Fritz Zwicky reported that galaxies in the Coma cluster moved too fast to be gravitationally bound to the visible matter. The case for “missing matter” solidified in the 1970s when astronomer Vera Rubin found a similar result, this time for stars in the outer reaches of individual spiral galaxies. Today, a rich and compelling body of astrophysical evidence has revealed not only that ∼85% of matter is invisible, but also that DM is “non-baryonic” meaning that it does not interact with light and falls outside of the Standard Model.

This astonishing discovery has triggered a global effort to uncover the nature of DM. Despite decades of experimental searches, however, no experiment has produced a convincing detection of DM. Although there is no guarantee that DM interacts with regular matter other than via gravity, a rich set of hypothesized interactions has been proposed, and associated experiments developed to search for DM via those interactions. Detecting and characterizing DM remains one of the biggest open questions in physics and is the subject of this Resource Letter.

With relatively few constraints on the nature of DM, the theory space for the physics of DM is exceptionally broad, with the mass of proposed DM candidates spanning at least fifty orders of magnitude. Experiments, meanwhile, tend to be sensitive to a restricted set of candidate species and masses, making exhaustive exploration of theoretical candidates challenging.

To date, experimental searches have mainly focused on two DM candidates: the weakly interacting massive particle (WIMP) and the axion, though a broader scope of theoretical candidates and search strategies are under development. Of particular interest are technologies from the world of quantum sensing that may open new windows in the search for DM.

Here, we provide an overview of the robust suite of astrophysical evidence for DM, the current theoretical frameworks for DM candidates, and the status of experimental searches, with a focus on direct detection experiments.

The US particle physics community engages in a long-term scientific study called the Particle Physics Community Planning Exercise (a.k.a. “Snowmass”) to identify important scientific questions, priorities, and opportunities. The most recent study, Snowmass 2021, included a detailed review of the status of DM theory and experiments. Reports from that study were released in September 2022 and provide a comprehensive overview of the current state of the field of DM research, as well as the desired trajectory for the coming decade. In the context of Snowmass, DM research falls primarily under the so-called “Cosmic Frontier.” Topical Groups within the Cosmic Frontier focus on different types of DM candidates: “particle-like” which includes the WIMP and “wave-like,” which includes the axion.

In addition to the Snowmass reports, many DM review articles and textbooks are available. Some are listed here; others are included in subsequent sections of this Resource Letter.

A robust suite of astrophysical observations strongly supports the existence of DM. Initially dubbed “missing mass,” evidence now suggests not only that the majority of the matter in the universe is non-luminous, but also that it is non-baryonic (defined in Sec. III B) and falls outside of the Standard Model of particle physics.

Fritz Zwicky's measurements of the radial velocities of individual galaxies within the Coma cluster 6 million light years away indicated that the galaxies were moving too rapidly to be bound by their mutual gravitational attraction. Zwicky assumed that the cluster geometry was statistically stable and used the Virial theorem to estimate the total gravitating mass of the cluster from the measured speeds of the galaxies, finding it to be 160 times larger than inferred from the luminosities of the galaxies. Zwicky interpreted this as evidence for the presence of non-luminous matter. We now know that measurements at that time were not sensitive to the x-ray emitting gas of the intracluster medium (ICM). Even after accounting for the ICM, however, the deficit persists. More recent observations show that about 1% of the mass is in stars (themselves within galaxies), 14% in the ICM, and 85% in DM, and that this is typical of rich galaxy clusters. Clusters are the largest gravitationally bound systems in the universe and so their composition is expected to be representative of the universe as a whole.

  • 14.

    Die Rotverschiebung von extragalaktischen Nebeln
    , Helv. Phys. Acta, 6, 110
    English and Spanish Translation of Zwicky's (1933) The Redshift of Extragalactic Nebulae
    ,” H. Andernach and
    , arXiv:1711.01693 (2017)

  • 15.

    The baryon content of galaxy clusters: A challenge to cosmological orthodoxy
    S. D. M.
    J. F.
    A. E.
    , and
    C. S.
    , Nature 366, 429

Vera Rubin and her colleagues found a similar result when they measured the orbital speed of matter in spiral galaxies as a function of the distance, r, from the galactic center—a so-called rotation curve. The gravitational influence of the visible matter in the spiral arms of the galaxies meant that the rotational speed of objects should decrease like r−1/2 at large radii. Instead, they found that the orbital speed was approximately independent of r (i.e., a “flat rotation curve”), exceeding predictions based on the amount of visible gravitating matter present in the galaxy.

Flat rotation curves are common in spiral galaxies and can be explained by a “halo” of DM that dominates the galaxy mass at large radius. The Standard Halo Model assumes an isotropic, isothermal sphere of DM with a density proportional to r−2, and radius O(10) times larger than the spiral arms of the galaxy. Measurements of our own Milky Way galaxy, which can be done with a home-made radio telescope (a great project for high school or undergraduate students), also reveal a flat rotation curve.

Zwicky's and Rubin's observations support the existence of non-luminous matter, but their data could not distinguish between (1) matter that escaped detection because the light it emitted was either too dim or at wavelengths that telescopes were not sensitive to, and (2) a fundamentally new kind of matter, beyond the Standard Model, that did not interact with light. We will use the term “baryonic” to refer to matter of the first kind, and “non-baryonic” in reference to the second. Astrophysical and cosmological observations indicate that DM is non-baryonic.

Take, for example, measurements of the anisotropy of the Cosmic Microwave Background (CMB). The temperature of the early universe was high enough to prevent neutral atoms from forming. At that time, photons interacted with the electrically charged plasma, and the mean free path of photons was very small. As the universe expanded and cooled, electrons bound with protons to form neutral hydrogen, causing photons to decouple from matter. The temperature and polarization of those photons encoded information about the structure of matter at the time of last scattering, and can distinguish between baryonic matter, which interacted with those photons, and non-baryonic matter, which did not. Precision CMB observations show that the ratio of the mass of baryonic matter to total matter (baryonic and non-baryonic) is 0.17, meaning that there is approximately 5 times as much DM as baryonic matter.

Further support for the non-baryonic DM paradigm comes from the observed abundance of light elements and the distribution of matter on the largest scales in the universe. In the hot early universe, nuclear reactions dictated the composition of the universe in a process called Primordial Nucleosynthesis or Big Bang Nucleosynthesis (BBN). BBN predicts that the relative abundances of light elements (deuterium, helium and lithium), depend strongly on the density of baryons (neutrons and protons). Measurements of the abundance of light elements therefore constrain the baryon density and are consistent with the CMB results (the baryonic density is five times smaller than the non-baryonic density). Any DM candidate must respect the BBN and CMB constraints and must therefore be “collisionless,” i.e., couples weakly to the baryons.

Measurements of large-scale structure in the universe provide further evidence that DM is collisionless. Matter that supports acoustic waves (i.e., baryonic matter) will develop density fluctuations during gravitational collapse due to the interplay between infalling matter and outward radiation pressure from photons. These fluctuations would manifest as characteristic oscillatory features in the matter density on cosmological length scales. Yet measurements of the large-scale structure power spectrum show no such features, adding still more support to the non-baryonic DM paradigm.

In sum, astrophysical observations tell us that non-baryonic DM exists over a range of spatial scales from small (galactic) to large (cosmological horizon), and in both the early universe and present time. DM moves at non-relativistic speeds (“cold DM”), otherwise small-scale structures in the universe would have been washed out. DM dominates the mass budget of the universe, is concentrated in galactic halos and clusters of galaxies, and is electrically neutral (or nearly so). The Standard Halo Model (SHM) posits a large, spherical DM halo with the stellar disk rotating relative to the halo, though recent data from the Gaia satellite indicate the presence of a radially anisotropic DM component as well. The local mass density of DM is measured to be ρDM ≈ 0.2–0.4 GeV/c2/cm3 (the mass of a proton in GeV/c2 is approximately 1). Our Solar System, which orbits the galactic center, is expected to move relative to the DM halo, thereby creating a DM “headwind” for an Earth-bound observer.

For the purposes of this work, we assume that DM exists, as evidenced by the robust observational data presented in Sec. III. An alternative explanation, however, is that our understanding of gravitational dynamics is incomplete. Several alternative paradigms, including Modified Newtonian Dynamics (MOND), aim to explain the astrophysical evidence in that context. However no modified gravity theory can yet explain the full suite of observational evidence for DM.

The extensive observational evidence for DM presented in Sec. III rules out all Standard Model particles, but otherwise provides only loose constraints on the nature of DM. DM candidates have been proposed with masses spanning 50 orders of magnitude, from 10−22 to 1028 eV/c2. Some DM models propose a single new particle to add to the Standard Model, while others postulate an entire dark sector with multiple DM states and interactions between them.

Two classes of DM candidates have commanded most of the attention of both theorists and experimentalists: Weakly interacting massive particles (WIMPs) and axions. In recent years, however, DM theory space and search strategies have significantly expanded to explore a broader range of candidates.

The term WIMP is used with different meanings in the literature. Here, we take a WIMP to be a particle of mass 10 GeV/c2–10 TeV/c2 that interacts via the weak force. A great deal of attention in both theory and experiment has been paid to WIMPs because of several notable properties. First, theories such as supersymmetry (SUSY) that attempt to explain the gauge hierarchy problem (i.e., why the weak force is 1024 times stronger than the gravitational interaction) also predict a WIMP-like particle. Second, WIMPs could be produced with the correct relic density in the early universe via thermal freezeout (the “WIMP miracle”). The WIMP, therefore, has the favorable property that it may solve two seemingly unrelated open questions in physics.

The thermal WIMP paradigm described above can be extended in many ways, while still reproducing the correct relic abundance. For example, scalar WIMPs have masses below 1 GeV/c2, while “asymmetric DM” posits an asymmetry between the DM particle and its antiparticle. Another approach assumes an entire “hidden sector” with its own matter and interactions. The hidden sector interacts with Standard Model particles via a “portal.” Still other models postulate non-thermal production of WIMP-like particles. A category called feebly interacting massive particles (FIMPs) have such small couplings to the Standard Model that they do not thermalize in the early universe but rather are produced by the decay of Standard Model particles.

In addition to WIMP DM (and its extensions), other forms of particle DM have been proposed. For example, the sterile neutrino, a hypothetical particle that mixes with Standard Model neutrinos but otherwise does not participate in the electroweak interactions, is a DM candidate with mass in the keV/c2 to MeV/c2 range. There are also a broad set of DM candidates called “ultra-heavy dark matter” (UHDM), whose masses range from ∼10 TeV/c2 up to the Planck mass,1019 GeV/c2.

The many proposed models of particle DM listed above, with masses ranging from O(eV/c2) to O(TeV/c2) and with a host of production mechanisms, are indicative of the lack of stringent constraints on DM properties. As explained next, DM may not even be particle-like.

For very low masses (m < 1 eV/c2), DM is better characterized as a wave than as a particle. For example, galactic DM with a mass of 0.2 μeV/c2 has a Compton wavelength of ∼1 km. Observations of the localization of DM in dwarf galaxies (length scales on the order of a kiloparsec for the smallest observed dwarf galaxies), combined with the uncertainty principle, provides a lower limit of m > 10−22 eV/c2 on the DM mass (otherwise the DM would not be localized in the galaxy). If m ≤ 100 eV/c2 the DM mode occupation number must be greater than unity to have enough DM to form the dwarf galaxy's potential, so Pauli's exclusion principle indicates that DM with mass in the range 10−22–102 eV/c2 must be bosonic. The parameter space of wavelike DM is relatively unexplored.

The canonical wave-like DM candidate is the so-called quantum chromodynamics (QCD) axion, with masses between 10−12 and 1 eV/c2 not yet ruled out by astrophysical constraints. The QCD axion arises in theories that attempt to explain the strong-CP problem in particle physics. In the Standard Model, the strong force can violate charge conjugation parity (CP) symmetry. The amount of CP violation is quantified by a parameter θ, which is constrained experimentally to be less than 10−10. To explain this small, but nonzero value of θ, the Peccei-Quinn mechanism promotes θ from a parameter to a dynamical variable that relaxes over cosmic time to its current value. Excitations of θ correspond to a particle, dubbed the “axion” that, generically, is “cold” (as in CDM) and has the appropriate relic abundance. The axion coupling strength to Standard Model fields is directly proportional to the axion mass, and so there is a bounded area of parameter space to search in for the QCD axion, meaning that an exhaustive search is possible in principle.

In addition to the QCD axion, axion-like DM and other wavelike DM candidates have been proposed as well. Axion-like particles (ALPs) share features with the QCD axion (a pseudo-scalar boson with small coupling and low mass), but do not solve the strong CP problem. They arise frequently in extensions to the Standard Model. ALPs have a wider range of possible masses and couplings than the QCD axion.

Many other DM candidates have been proposed beyond the several we have discussed so far, each with their own origin story. Primordial black holes (PBHs) and superfluid DM are two examples that have attracted recent attention. PBHs formed from density fluctuations in the early universe. They are long-lived (longer than the age of the universe if their mass is greater than 10−19 M), and current constraints admit several mass ranges in which PBHs could either constitute all of the DM or contribute appreciably to it.

Superfluid DM, on the other hand, provides a unified framework for DM and MOND. On cosmological scales, the DM forms a cold, collisionless fluid. As the DM coalesces via gravitational collapse, it undergoes a phase transition to a superfluid that experiences both Newtonian and non-Newtonian couplings to surrounding baryons. This non-Newtonian interaction can match the empirical MOND law.

Searches for DM may take many forms including direct detection, indirect detection, collider searches, and cosmological probes. Direct detection experiments seek evidence for the interaction of a DM particle with a target in the laboratory. Signatures of these interactions depend on the type of DM and detector but may include the production of charge (ionization), light (scintillation), and heat (thermal changes), phase changes, or resonant couplings, to name a few. A challenge of direct detection is the discrimination between the rare (and typically small) signal of interest and abundant backgrounds that mimic DM interactions.

Indirect DM detection searches seek evidence for Standard Model particles (e.g., photons, neutrinos, protons) that are produced in DM decays or annihilation. For example, observations of gamma rays from the center of our galaxy reveal an as-yet unexplained excess that peaks in the 1–3 GeV energy range. One possible interpretation is that the excess is due to the annihilation of DM with mass < 100 GeV/c2. But other (more mundane) astrophysical processes could be the source (e.g., gamma ray emission from pulsars). This example illustrates a fundamental challenge of indirect detection – distinguishing between DM and non-DM origins for astrophysical signals of interest.

Particle colliders, such as the Large Hadron Collider, could produce DM particles if the energy available in the collision is sufficiently large (exceeds the rest energy of the DM particle). Although DM produced at a collider would be unlikely to interact in the detectors surrounding the collision point, DM production could be inferred via precision measurements of the resonance widths of particles that couple to DM, or missing transverse momentum in a collision which could be attributed to a particle that escaped detection (e.g., a DM particle, but also neutrinos). Challenges with these searches include detector “pile-up” in which multiple collider interactions happen at nearly the same time and are erroneously grouped together during analysis, and non-collision backgrounds, such as cosmic rays whose transverse momentum may be mistakenly associated with a collision event.

Cosmological probes allow for an exploration of DM across enormous ranges of space and time. The interaction of DM with other particles in the early universe can alter the abundance of elements and affect the formation of large-scale structure. Consequently, measurements of those quantities can provide constraints on DM candidates and their properties that complement the results from the other search strategies mentioned above.

The rest of this Resource Letter focuses on the status and prospects of direct detection experiments that look for evidence of the interaction of DM in a laboratory detector.

Particle-like DM direct detection experiments take many forms. For example, traditional WIMP searches seek evidence for the interaction between the WIMP and a target nucleus in the detector. For typical WIMP masses (102±1 GeV/c2), and WIMP speeds in the galactic halo (v/c ∼ 10−3), the interaction is non-relativistic and produces a nuclear recoil with kinetic energy O(10 keV). Detecting that recoil and distinguishing it from other non-WIMP energy depositions in the detector is very challenging. Great progress in this field has been made, however, with detector sensitivities increasing by over five orders of magnitude in the past two decades.

WIMP-nucleus interactions may be spin-dependent or spin-independent, with a coherent enhancement favoring nuclei with a large atomic mass number for spin-independent interactions. Observable signatures of a WIMP-induced nuclear recoil include ionization (charge), scintillation (light), thermal deposition (heat), or phase changes (bubble nucleation), and a detector may be sensitive to one or more of these signatures.

Background events can contaminate the WIMP signal of interest, and direct detection experiments pay careful attention to both suppressing and tagging backgrounds. To reduce the background rate, detectors operate underground, and are typically encased in shielding (usually multiple, nested, layers of shielding made of different materials such as water, lead, copper). Furthermore, the detector components themselves must be radiopure so as not to be a significant background source. Some detector target materials (e.g., liquid xenon) provide significant self-shielding, in which background particles interact in the outer portions of the sensitive volume, leaving a lower-background inner “fiducial” region with which to carry out the WIMP search.

In addition to reducing the number of background interactions, experiments can leverage differences in the features of signal-like and background-like events to tag and reject remaining backgrounds. For example, the time distribution of scintillation light in liquid argon is different for nuclear recoils (WIMP-like events) vs. electron recoils (backgrounds). In liquid xenon, the ratio of scintillation light to ionization charge in an event (the so-called S1 vs. S2) is a signal/background discriminant, and in cryogenic silicon and germanium detectors, the ratio of the ionization and phonon yields can be used to separate nuclear recoils from electron recoils.

The current leading spin-independent limits were set by experiments that use liquid xenon (LXe – e.g., the LZ, XENONnT, PandaX experiments) or liquid argon (LAr – e.g., DarkSide). Frontier spin-dependent limits come from a broader suite of experiments including LXe detectors as well as fluorine-rich bubble chambers, germanium sensors, and scintillating crystals. Currently operating detectors that use LXe, germanium/silicon sensors, and scintillating bubble chambers have the leading projected sensitivities, while many new detection techniques have been proposed including the use of supercooled water as a target medium.

Coherent scattering of neutrinos in WIMP detectors will eventually become a source of background. This so-called “neutrino fog” may limit the sensitivity of these devices as the neutrinos cannot be shielded, though differences in the signal distributions (energy and direction for example) may allow experiments to statistically distinguish WIMP-like from neutrino-like events.

A tell-tale signature of galactic halo DM is the order-unity asymmetry in the direction of WIMP-induced nuclear recoils. The motion of the Earth through the halo produces a WIMP headwind. The nuclear recoils should therefore cluster in the opposite direction. A WIMP detector with directional sensitivity could measure this angular asymmetry and make a convincing detection of halo DM. Only a handful of events are required because the WIMP angular distribution is significantly different from all known backgrounds. Furthermore, a directional detector could also see through the neutrino fog because of differences between the directions of neutrino-induced and DM-induced recoils.

Many technical challenges stand in the way of the design and operation of a large-scale direction-sensitive WIMP detector, however. In solid and liquid targets, for example, low-energy nuclear recoils travel a very short distance (sub-micron). This makes the reliable reconstruction of recoil tracks challenging, though some progress has been made using fine-grained emulsions. In the absence of geometric track reconstruction, it may be possible to infer the recoil direction via a directional dependence in the detector response (e.g., anisotropic light yield in crystals). Quantum sensing techniques may also be able to infer the nuclear recoil path based on the crystal lattice damage it left behind. In low-pressure gas targets, the recoil track is longer (millimeters), and directional sensitivity has been demonstrated. The CYGNUS proto-collaboration is pursuing large-scale gas detectors for directional dark matter detection.

Direct detection efforts for particle-like DM have largely focused on WIMPs. But direct searches for light DM (< 1 GeV/c2) and UHDM (> 10 TeV/c2) are areas of growing interest, with novel detection strategies under development. Light DM can interact with target electrons or nucleons, and pioneering light DM searches were based on the novel analyses of data from traditional WIMP detectors that enabled lower energy threshold (albeit at the expense of a higher background contamination). The Migdal effect can also be leveraged to achieve lower energy thresholds. At masses below ∼ 1 MeV/c2, the DM could, for example, interact with collective modes in the detector (e.g., phonons).

UHDM, on the other hand, would have a low flux at the Earth (about 1 event per square meter per year for a particle with the Planck mass), but will suffer little deflection when traversing a detector. Multiple interactions would create a linear track that would directly reveal the DM direction. A complementary approach to UHDM detection uses ancient minerals extracted from ∼ 10 km below the surface of the Earth. This so-called paleo-detection strategy benefits from an extremely long exposure time (109 years). Alternatively, UHDM may interact via a long-range force, in which case a coherent interaction with a macroscopic target may produce a measurable signal (e.g., in an optically levitated nanogram mass).

One direct detection group, the Dark Matter (DAMA) experiment (subsequently DAMA/LIBRA), has claimed to have detected DM. The collaboration reports an annual modulation in the detected event rate since 1995, and attributes it to the motion of the Earth through the DM WIMP halo. Indeed, the motion of the Earth and Sun through the DM halo of the Milky Way would produce a head-wind of WIMP that increases the event rate when the Earth's velocity adds constructively with the Sun's (in the galactic frame), and decreases the rate when the velocity vectors add destructively. The period and phase of the modulation measured by DAMA matches expectations.

This detection claim has been met with skepticism, however, not only because of the absence of signals in other experiments with adequate sensitivity to the DM mass and interaction cross section indicated by DAMA, but also because of the possibility that low-energy events in the detector could be attributable to unmodeled backgrounds. This has led to efforts to reproduce the DAMA result using the same detector target material as DAMA (thallium-doped sodium iodide crystals). Two such experiments, COSINE-100 and ANAIS-112, have released their initial results, and neither sees the DAMA excess after three years of operation. There is no question that DAMA sees a modulated event rate, but the claimed DM origin of the modulation is a topic of active debate.

Much like for particle DM, the search for wave-like DM requires a suite of different experiments, each with complementary technologies and sensitivities. In parallel with the large-scale and well-established QCD axion searches, many new techniques are currently under development—at the R&D or small prototype scale—many of which leverage techniques from the world of quantum sensing.

Wave-like DM searches can be grouped into three broad categories, named by the source of the DM that they seek: Haloscope, helioscope, and laboratory experiments. Haloscopes seek to measure the interaction of Galactic halo DM. Helioscopes, on the other hand, specifically target axions or axion-like particles produced in the Sun. Finally, lab experiments seek to produce, and subsequently detect, axions or axion-like particles in the laboratory.

The QCD axion search community is advocating a “definitive axion search program” to cover the full region of axion parameter space (axion coupling strength vs. mass) from 10−12 to 1 eV/c2 in the coming two decades. Masses below that range have been ruled out by constraints from spinning black holes, while masses above that range are constrained by stellar astrophysics. Traditional QCD axion experiments search for the conversion of the axion into photons in the presence of a strong magnetic field within a resonant cavity via the inverse Primakoff effect. By scanning the resonant frequency of the cavity fcavity, the experiment scans across a range of axion masses (fcavitymaxion). For example, the Axion Dark Matter Experiment (ADMX), a leading axion search experiment, aims to cover the mass range of 2–8 μeV/c2 via mechanical tuning of the resonant cavity frequency.

The rate that a resonant cavity search experiment can scan axion mass scales as QV2C2B4T−2 with Q the resonant cavity quality factor, V the cavity volume, C the geometric form factor that characterizes the coupling of the axion to a particular cavity mode, B the magnetic field strength, and T the noise temperature of the readout. To maintain good axion-mode coupling (C) the cavity size should scale with the axion's Compton wavelength ∝ 1/maxion. At high axion mass, smaller detection volumes are therefore indicated (V ∝ 1/maxion3), but that decreases the mass scan rate (∝ V2) rapidly with axion mass. To address this challenge, experiments could scale up by replicating many small detection volumes, boosting the magnetic field strength (e.g., high-field superconducting magnets, and rare-earth superconductors), and raising the cavity resonant factor (superconducting films or low-loss dielectrics). Additionally, the axion scan rate can be enhanced by reducing readout noise below the standard quantum limit (e.g., using squeezed states or single-photon counting). For example, the HAYSTAC experiment (Haloscope at Yale Sensitive to Axion cold dark matter), which targets the axion mass range of 15–50 μeV/c2, has demonstrated a quantum squeezed state receiver that enhances the mass scan rate by a factor of 2–3. Other proposed readouts seek to measure only the amplitude of the signal photon, thereby relegating the quantum noise to the unmeasured phase. The use of qubit-based quantum non-demolition measurement can also enhance the axion sensitivity.

At lower axion masses (below the μeV/c2 scale), corresponding to lower resonant frequencies, it becomes impractical to build cavities large enough for a resonant cavity search (the axion's Compton wavelength is too large). An alternative approach uses a toroidal magnet in which the axion field induces an electric current that can be sensed by a superconducting coil or lumped element resonators.

If axions exist, then they would be produced in abundance in the cores of stars where plasma photons convert to axions via the Primakoff process. In the Sun, the expected axion energy is a few keV. Solar axions arriving at the earth could then convert to x-rays in a strong magnetic field. An axion helioscope uses a long, high-field magnet pointed at the Sun to search for this signal. Experiments such as CAST (CERN Axion Search Telescope) have ruled out large regions of ALP parameter space. The projected sensitivities of next-generation helioscopes such as BabyIAXO and IAXO reach the QCD axion.

Rather than seeking axions from the galactic halo or the Sun, some groups, such as The Any Light Particle Search II (ALPS II) and Optical Search for QED Vacuum Birefringence, Axions, and Photon Regeneration (OSQAR), aim to produce axions in the laboratory and subsequently detect them. These laboratory experiments do not suffer from astrophysical uncertainties, but they are challenging because both the axion creation process and conversion back to photons are rare. Sometimes called “light-shining-through-wall” experiments, photons from a powerful laser in a strong magnetic field could produce an axion that would then travel through a light-blocking wall and re-convert to photons on the other side.

The axion and axion-like particle experiments described above seek evidence for the electromagnetic coupling of the axion. Axion and ALPs could also couple to fermions or to the nuclear electric dipole moment (EDM). Fermionic coupling can be probed using magnetometers (among other approaches), while nuclear magnetic resonance spectroscopy and storage ring methods can probe EDM couplings.

The mystery of DM is one of the most significant scientific questions of our time. Astrophysical observations provide strong evidence that non-baryonic, electrically neutral, long-lived DM makes up 85% of the matter in the universe. The parameter space of theoretical DM candidates remains expansive. Impressive progress has already been made in theoretical framework development and experimental searches, and the coming decade will undoubtedly see exciting progress on both fronts. DM has not been directly detected (at least not convincingly), but many DM direct detection experiments that are currently in operation will achieve at least an order of magnitude improvement in sensitivity, and therefore provide legitimate discovery potential. Meanwhile, ongoing R&D into novel DM detection techniques will open new windows of exploration in the hunt to uncover the physics of DM.

This work is supported by the Gordon and Betty Moore Foundation through Grant GBMF11565 and Grant DOI https://doi.org/10.37807/GBMF11565.